Maternal serum thyroid-stimulating hormone (TSH) has been estimated in 38 pregnant women with three new specific nonradioactive monoclonal immunoassays which can measure very low concentrations. Values for the Pharmacia-LKB 'DELFIA' fluoroimmunoassay and for the Amersham 'Amerlite' luminescence immunoassay were statistically identical over the range 0.15-1.5 uIU/ml WHO 2nd IRP 80/558, but thereafter the Amersham values were slightly lower. The Abbott 'IMx' assay system gave slightly higher results commensurate with the quoted higher non-pregnant normal values. The Amersham assay was preferred on practical grounds. There was no relationship of TSH levels to HCG levels, or to the length of pregnancy. Four women had apparently zero TSH levels by the Amersham and Pharmacia-LKB assays, although the Abbott assay could detect very low concentrations. Together with women whose TSH levels were below the normal non-pregnant range for each assay, there were a total of eight women (21%) with TSH levels below 'normal'. This suppression of maternal pituitary TSH levels during pregnancy was considered to be due to the central feedback inhibitory thyrotrophic activity of HCG, as a separate placental thyrotrophic hormone is believed not to exist. The enlargement of the maternal thyroid gland and the increased production of thyroid hormones during normal pregnancy is likewise to be attributed to HCG rather than to an increase in TSH production.